When quantum mechanics predicts the outcome of a measurement it rarely predicts a single definite
value. Instead it gives a probability distribution over the eigenvalues of the observable. The
standard deviationΔA quantifies the spread of that distribution — it is the quantity
that appears directly in the Heisenberg uncertainty relation ΔxΔp≥ℏ/2.
Expectation value
Let A be an observable (a Hermitian operator) and ∣ψ⟩ a normalised state. If A has
eigenvalues a1,a2,… with corresponding probabilities p1,p2,… in the state
∣ψ⟩, the expectation value is
⟨A⟩=∑ipiai.
This is precisely the weighted average of the possible measurement results.
Standard deviation (uncertainty)
The uncertaintyΔA is the root-mean-square deviation from the mean:
(ΔA)2=⟨(A−⟨A⟩)2⟩=⟨A2⟩−⟨A⟩2.
Both identities are equivalent; the second is usually faster to calculate. In practice you compute
⟨A2⟩=∑ipiai2 and then subtract ⟨A⟩2.
Worked example
Suppose A has two eigenvalues, a1=+1 (probability p1=3/4) and a2=−1
(probability p2=1/4).
Step 1 — expectation value:
⟨A⟩=43(+1)+41(−1)=43−41=21.
Step 2 — expectation value of A2:
Since a12=a22=1,
⟨A2⟩=43(1)+41(1)=1.
Step 3 — standard deviation:
(ΔA)2=⟨A2⟩−⟨A⟩2=1−41=43,
ΔA=43=23≈0.866.
Why it matters
The Heisenberg uncertainty principle bounds the product of two such standard deviations for
non-commuting observables. Mastering how to compute ΔA for a single observable is the
essential prerequisite for understanding that principle quantitatively.
Try it
This is a numerical exercise — your code should return a number, not a circuit.
An observable A has eigenvalues +1 (probability 3/4) and −1 (probability 1/4). Compute
ΔA=⟨A2⟩−⟨A⟩2.
Run your code to see the quantum state.
Sign in on the full site to ask questions and join the discussion.